quality-indicators-2017

More documents

Recommendations

Info

QUALITY INDICATORS <strong>2017</strong> 3.13.2 Calculation procedure If the announced date of the first release equals value equals PT 3 D D . 1 N D N , and the actual date of the first release is D 1 , the indicator 3.13.3 Explanation and calculation examples Indicator values are always calculated and expressed in terms of the number of days. If the first results are actually released prior to the announced date of release of the first results, the indicator may also have a negative value. Example 1: If the announced date of first release is 16 March 2009 and the results were published on 20 March 2009, the indicator value is equal to PT 3 (20.3.2009) (16.3.2009) 4 . Example 2: If the announced date of the first release is 16 March 2009 and the results were published on 13 March 2009, the indicator value is equal to PT 3 (13.3.2009) (16.3.2009) 3 . 3.14 Length of comparable time series (P1) 3.14.1 Description Indicator P1 expresses the length of time series from the last break in the time series, or the number of time points in the time series from the last break. The indicator value refers to key statistical surveys. 3.14.2 Calculation procedure If T is the last reference period for which results have been released (the last point in the time series) and T 0 is the first period with statistical results that are still comparable (the first time point after a possible break), the indicator value is determined as follows: P1 T T 0 . If the time series has no breaks, the indicator value expresses the length of the entire time series. 3.14.3 Explanation and calculation examples The indicator value expresses the number of periods in the time series; this means that the measurement unit depends on the periodicity of the survey. In a monthly survey, the value of the indicator is thus expressed as the number of months, while in a quarterly survey it is expressed as the number of quarters, etc. The value of the indicator is to a great extent determined by the definition of the break in the time series. The fact is that it is impossible to establish exact criteria by which the break would be identified, which means that usually – at least partly – a subjective assessment is required. In general, we could say that a break in a time series occurs when one or several aspects of the survey change (e.g. in the source of data or the methodology used) and then the final results are no longer comparable. The OECD glossary of statistical terms defines in its original text time series breaks as follows: Breaks in statistical time series occur when there is a change in the standards for defining and observing a variable over time. Such changes may be the result of a single change or a combination of multiple changes at any time point of observation of the variable. Example 1: Let us assume that the reference period January 2002 represents the beginning of a comparable time series of a monthly survey. Then, the value of the indicator for the reference period June 2009 is P 1 ( June09 January02) 90 (months). 23
QUALITY INDICATORS <strong>2017</strong> Example 2: If in a yearly survey the year of the first comparable release of results is 1998, upon the release of the results for 2009 the indicator value is equal to P 1 12 (years). 3.15 Coherence between provisional and final data (S1) 3.15.1 Description The absolute and the relative differences between the value of statistics upon the first release and the value upon the release of the final data. As revisions, only those corrections are considered which are part of the regular procedure for the repeated releases of statistical results. Corrections due to discovered errors do not apply to this category and are not considered in the calculation of the indicator. Even if the revision policy determines several versions of (provisional) data for the same reference period, for reason of simplicity, only the difference between the value of the first and the value of final release are calculated. The indicator value refers to key statistics. 3.15.2 Calculation procedure Let us say that X z is the first and X k the final value of the released statistical result (statistic). The calculation of the indicator is classified with regard to the type of statistical result. If the released result is a "dimensionless" statistic (e.g. rate, index), the indicator value is calculated as the absolute difference: S 1 . X k X z If the released result is a "dimensional" statistic (e.g. sum, average), the indicator value is calculated as the relative difference: S X X k z 1 . X k 3.15.3 Explanation and calculation examples Instead of the title coherence between the final and provisional results, the indicator could also be referred to as coherence between the most recently released results and the provisional results. Since we usually calculate the value of indicator S1 after the release of the final results, the most recently released results are also the final results. In particular in short-term statistics, the results for the same reference period are released several times (sometimes also more than 10 times). In such cases, at each release of results the difference compared to the first released result could be calculated. In order to simplify the calculation and to provide clarity, the indicator is only defined as the (relative) difference between the first and final releases. In such cases, a supplementary indicator can also be calculated and released, for example, the maximum deviation of the last release from the first release, which would be calculated as max X m i X z , where m is the number of releases of statistical i 1 results (for the same reference period) after the first release of results. Example 1: In this case, (hypothetical) first and final results in the monthly survey are stated for all months in 2008. Since the results are indices, the indicator value is calculated as the absolute difference. The average (annual) value of the indicator is also given. 24
Page 1: METHODOLOGICAL MANUALS QUALITY INDI
Page 4: QUALITY INDICATORS 2017 FOREWORD On
Page 7 and 8: QUALITY INDICATORS 2017 3.12.2 Calc
Page 9 and 10: QUALITY INDICATORS 2017 2 GENERAL I
Page 11 and 12: QUALITY INDICATORS 2017 Below follo
Page 13 and 14: QUALITY INDICATORS 2017 The value o
Page 15 and 16: QUALITY INDICATORS 2017 3.5 Unit no
Page 17 and 18: QUALITY INDICATORS 2017 which no da
Page 19 and 20: QUALITY INDICATORS 2017 3.7.2 Calcu
Page 21 and 22: QUALITY INDICATORS 2017 If w is the
Page 23: QUALITY INDICATORS 2017 3.11.2 Calc
Page 27 and 28: QUALITY INDICATORS 2017 Example 1:
Page 29 and 30: QUALITY INDICATORS 2017 4.1.2 Overc
Page 31 and 32: QUALITY INDICATORS 2017 Figure 5.2:
Page 33 and 34: QUALITY INDICATORS 2017 6 AUTOMATIS
Page 35 and 36: QUALITY INDICATORS 2017 7 STANDARD
Page 37 and 38: QUALITY INDICATORS 2017 8 SUMMARY O
Page 39 and 40: QUALITY INDICATORS 2017 Seljak, R.
Page 41 and 42: QUALITY INDICATORS 2017 T2. Time la
Page 43: QUALITY INDICATORS 2017 Code list w

quality-indicators-2017

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?